Fotor Skills

Current public skill version: 1.0.22

This repository stores reusable Agent Skills for Fotor AI.

Official product page: https://developers.fotor.com/fotor-skills/

Version Management

• Keep skills/fotor-skills/SKILL.md top-level version aligned with the current published skill version.
• Keep CHANGELOG.md updated for GitHub / npx skills installs.
• Avoid a separate root-level VERSION file.

Current Skill

fotor-skills

An all-in-one AI photo editor and AI video generator workflow for generating, editing, transforming, and enhancing images and videos with the fotor-sdk package.

For API key application and product details, see https://developers.fotor.com/fotor-skills/.

SDK compatibility follows the latest fotor-sdk release installed by scripts/ensure_sdk.py.

It supports:

• Text-to-Image (text2image)
• Image-to-Image (image2image)
• Image Upscale (image_upscale)
• Background Removal (background_remove)
• Text-to-Video (text2video)
• Single Image-to-Video (single_image2video)
• Start/End Frame Interpolation (start_end_frame2video)
• Multiple Image-to-Video (multiple_image2video)
• Credit lookup (get_credits_sync)

The skill includes setup scripts, execution tooling, model references, and parameter documentation.

Repository Structure

.
├── README.md
└── skills/
    └── fotor-skills/
        ├── SKILL.md
        ├── agents/
        ├── references/
        └── scripts/

Install

ClawHub

clawhub install fotor-skills

GitHub / npx skills

npx skills add https://github.com/fotor-ai/fotor-skills.git --skill fotor-skills --copy -y

Quick Start

1. Enter the skill directory:

cd skills/fotor-skills

2. Create a local virtual environment with uv:

uv python install 3.12
uv venv --python 3.12 .venv

3. Upgrade the SDK to the latest version:

./.venv/bin/python scripts/ensure_sdk.py

4. Configure your API key. Recommended local .env setup:

cat > .env <<'EOF'
FOTOR_OPENAPI_KEY=<your_api_key>
EOF

set -a && source .env && set +a

5. Run a sample task:

cat <<'EOF' | ./.venv/bin/python scripts/run_task.py
{"task_type":"text2image","params":{"prompt":"A cat astronaut","model_id":"seedream-4-5-251128"}}
EOF

Batch Execution

You can run multiple tasks in parallel:

cat <<'EOF' | ./.venv/bin/python scripts/run_task.py --concurrency 5
[
  {"task_type":"text2image","params":{"prompt":"A neon city","model_id":"seedream-4-5-251128"},"tag":"img-1"},
  {"task_type":"text2video","params":{"prompt":"A futuristic skyline","model_id":"kling-v3","duration":5},"tag":"vid-1"}
]
EOF

Input and Output Format

• Input: JSON object (single task) or JSON array (batch).
• Output: structured JSON with fields such as task_id, status, success, result_url, error, elapsed_seconds, creditsIncrement, and tag.

Scripts

• scripts/ensure_sdk.py
  • Install or upgrade fotor-sdk with uv.
• scripts/upload_image.py
  • Upload a local image file and return a reusable file_url.
• scripts/run_task.py
  • Run one or more OpenAPI tasks from JSON input.
• scripts/check_skill_update.py
  • Checks whether a newer version of the installed skill is available.

For credit lookup via client.get_credits_sync(), the SDK returns a dict like {"businessId": "", "total": 2000, "remaining": 1973}.

Model and Parameter References

Use these files when selecting models and building parameters:

• skills/fotor-skills/references/image_models.md
• skills/fotor-skills/references/video_models.md
• skills/fotor-skills/references/parameter_reference.md

Environment Variables

• FOTOR_OPENAPI_KEY (required): API key for authentication.
• FOTOR_OPENAPI_ENDPOINT (optional): API base URL. Defaults to https://api-b.fotor.com.

Common Issues

• FOTOR_OPENAPI_KEY not set
  • Add it to .env, source the file, then retry.
• Unknown task_type
  • Use one of the 8 supported task types listed above.
• Unsupported model_id or invalid parameters
  • Cross-check against references/ docs.

Contributing

When updating this repository:

• Keep skills/fotor-skills/SKILL.md top-level version aligned with the published version.
• Keep CHANGELOG.md aligned with the published version.
• Update reference documents when model lists or capabilities change.
• Keep runnable command examples in SKILL.md and this README aligned.